Polarimetric soil moisture retrieval using an iterative generalized hybrid decomposition technique

L-band fully polarimetric SAR data, acquired within several airborne measurement campaigns from 2006 to 2013 over Germany, are used for soil moisture estimation at high spatial resolution and under distinct agricultural vegetation cover. The fully polarimetric L-band data were acquired by DLR's E-SAR and F-SAR sensors in order to investigate their potential for soil moisture retrieval. A newly developed, iterative, generalized, hybrid decomposition and inversion approach is applied for a flexible soil moisture estimation under a temporally and spatially varying agricultural vegetation. Therefore an iteration-adaptive, generalized scattering model of the vegetation is used. First results of the inverted soil moisture under vegetation cover reveal high inversion rates for the agricultural test sites (>95%). The inverted soil moistures are validated with in situ measurements from simultaneously conducted field campaigns leading to a minimum root mean squarer error (RMSE) of 4.1vol.% for a variety of crop types and in situ moisture conditions (ranging from 5vol.% to 40vol.%).